AI assistants need
discipline, not just knowledge

A curated library of 41 executable skills and 9 workflows that teach AI coding tools how to plan, test, implement, document, and review Rails work using production-minded conventions.

Tests gate implementation
41 atomic skills
9 callable workflows

Generic code dumps without guardrails

Most AI coding assistants produce plausible-looking code that lacks tests, skips review, and ignores your team's conventions. The result: inconsistent quality, missed edge cases, and code that looks right but breaks in production.

  • Implementation written before tests exist
  • No checkpoint to confirm behavior boundaries
  • Self-review skipped or superficial
  • Different assistants produce different conventions
"Add a status column to orders..." One migration: add_column :orders, :status Production locks table, downtime ensues

Skills are executable instructions, not documentation

Each skill is a SKILL.md file containing task-specific instructions, conventions, and hard gates that an AI agent reads and follows. They are not tutorials to read once. They are constraints that run every time.

Atomic

One skill per concern. plan-tests chooses the first failing spec. write-tests writes it. code-review reviews it. No monolithic guides.

Gated

Hard gates enforce discipline. The Tests Gate means no implementation code can exist until the test exists, runs, and fails for the right reason.

Chained

Each skill names the next skill to load in its Integration table. write-tests → implement → write-yard-docscode-review. Skills are building blocks; workflows are the value.

The TDD Feature Loop

The most-used daily workflow. Every feature, bug fix, and refactor follows the same repeatable cycle. The agent cannot skip a checkpoint without your explicit approval.

1

load-context

Sync schema, routes, patterns

2

plan-tests + write-tests

Choose and write the first failing spec

Checkpoint: Test Feedback

You approve the test before any code is written

Checkpoint: Implementation Proposal

You approve the approach before implementation

3

Implement → Test passes → Refactor

Minimal code, green test, clean up

4

Docs → Review → PR

YARD, self-review, merge

41 skills, 9 workflows, 10 categories

Each skill has a SKILL.md entry point. Some have companion files for examples, test templates, patterns, or heuristics. Workflows chain skills into full development loops.

Planning

create-prd generate-tasks plan-tickets

Testing

plan-tests write-tests test-service triage-bug

Code Quality

code-review security-check refactor-code apply-code-conventions apply-stack-conventions

Architecture & DDD

define-domain-language review-domain-boundaries model-domain review-architecture

Rails Implementation

create-service-object implement-background-job implement-authorization implement-hotwire

APIs & Engines

implement-graphql integrate-api-client version-api create-engine test-engine

Workflows chain skills into full cycles

tdd
quality
review
engine
bug-fix
graphql
migration
background-job
setup

A safe migration request

See the difference between a naive request and a skill-guided request. The skill enforces a multi-step process that protects your production database.

!

Naive request

"Add a status column to orders with default 'pending' and not null."

AI generates:

add_column :orders, :status,
  :string, default: "pending",
  null: false
Production table lock. Downtime.

Skill-guided request

"Add a status column to orders. Follow review-migration."

AI generates:

# Step 1
add_column :orders, :status, :string
# Step 2: Backfill in background job
Order.where(status: nil).update_all(status: "pending")
# Step 3: Add constraint
change_column_null :orders, :status, false
No table lock. Safe deployment.

Three ways to invoke

Use MCP for autonomous tool calls, chat commands for explicit control, or the CLI for manual installation and evaluation.

1

MCP Server

The recommended path. The agent discovers skills through list_skills and loads them on demand via use_skill. Context stays small.

docker run --rm -i
igmarin/rails-agent-skills-mcp:5.1.5
2

Chat Commands

Force the agent to use a specific skill. In Cursor or Windsurf, prepend with @. In Claude Code, describe the goal and the agent loads the right skill.

"Follow the TDD workflow
for this task"
3

GitHub CLI

Install specific skills into your agent host. Pin to release tags for reproducible installs. Works with Claude, Cursor, Windsurf, and Codex.

gh skill install
igmarin/rails-agent-skills
plan-tests --scope user

Start with the TDD loop

The fastest way to experience the difference: pick a small feature, run load-context, then plan-tests. Watch the agent stop at checkpoints, confirm boundaries with you, and only then write code.

MCP Registry · skills.sh · Smithery · Glama · Tessl